Objective: To evaluate a newly developed biofeedback device enabling lower-limb amputees to identify various floor conditions.

Design: Self-control with repeated measures (with and without the biofeedback device) within the amputee group, and group control comparing between amputee and nonamputee groups.

Setting: University locomotion laboratory.

Participants: Five lower-limb amputees and 8 nonamputees (N=13).

Interventions: A wearable biofeedback device, which identified different floor conditions by analyzing the force patterns under the prosthetic feet and provided vibration cues in response to different floor conditions, was provided to the amputees.

Main Outcome Measures: The subjects stepped on a foam platform concealing a small object or no object at 1 of the 4 locations of the foot sole. Subjects were asked whether there was a small object under their feet and the location of the object if it existed. The test was repeated with 4 different object types and 4 object locations. The success rate of floor identification was evaluated.

Results: Without the biofeedback device, nonamputee subjects (76.56%) identified floor conditions better than amputees (22.5%) significantly (P<.001). On using the biofeedback device, the amputees significantly improved (P<.01) their success rate showing no significant difference (P=.746) compared with the nonamputees. No significant differences were found among object types (P=.689).

Conclusions: Amputees performed significantly worse than nonamputees in recognizing the different floor conditions used in this experiment. With the biofeedback device, amputees significantly improved their abilities in identifying different floor conditions. Future attempts could configure the device to allow it to provide warning signals in response to fall-inducing conditions.